Predicting Optimal Deep Brain Stimulation Parameters for Parkinson's Disease using Functional MRI and MSVM

Parkinson's disease (PD) is a progressive illness of the central nervous system primarily caused by neuronal degeneration in the substantia nigra of the brain. Blood uric acid level is an emerging biomarker for Parkinson's disease (PD). Despite numerous studies to the contrary, the relationship between Parkinson's disease, diabetes, and the outcomes of specific treatments remains unclear. A collection of machine learning (ML) models was created to predict Parkinson's disease based on MRI images. This work makes use of Parkinson's progressive Markers initiative (PPMI) dataset. Initially, the feature extraction process used VGG-16 and HOG for detecting Parkinson's disease. In following stage, a predicted classification of Parkinson's patients and healthy controls is produced using the Multi-class Support Vector machines (MSVM) model in an effort to enhance the final output of overall classification model. The individuals can use an accessible database created by the Parkinson's progression Markers Initiative (PPMI) to assess the implemented model. When compared to existing methods like Convolutional Neural Network with Fuzzy Rank Level Fusion (CNN-FRLF), and AlexNet with quantum transfer learning method, implemented method achieved high values of 99.53% accuracy.